Not decided. I have a couple of Crown amps with built-in DSP I can use for testing without having the NC500/DSP setup running so that will be first but it only offers a basic xo. I'm considering approaching Danville Signal for a custom unit that would also have driver correction but need to settle on a driver/enclosure first so I may just use the Crown amps for testing and optimize the design with DSP after this is done, then add the NC500s.
The only real downside of any of the SHARCs is that you are limited to closed, proprietary development tools.
Danville signal makes a user friendly platform. Besides if it get the job done, what's it matter? I like the 64 bit floating point processing. Also you can clock sync the chips with IEEE1588. That's a big problem for most of the DSP based units if you want to build the DSP into the speaker. Lack of clock sync.
It matters to me because I don't like to depend on tools and environments that depends on the continuing support from a commercial supplier that might not be in business in 2 years. I also like to be able to modify the software myself - this is a DIY forum, after all
It matters to me because I don't like to depend on tools and environments that depends on the continuing support from a commercial supplier that might not be in business in 2 years. I also like to be able to modify the software myself - this is a DIY forum, after all
They provide a platform that can be tweaked. Just like Hypex did for your unit.
So just like with the Hypex DSP tools, when the vendor stops supporting them, I will be screwed (unless I keep running a virtual machine under linux with an old version of Windows).
hypex ncore
Until the board fails, the programming tools will be just fine. There's so much functionality you won't need to update for eternity. Besides once you have it tuned to how you want it, why to you need to change it? Unless you're doing room correction. Either way who offers a better solution? Unless you go PC based with a multichannel DAC.
Until the board fails, the programming tools will be just fine. There's so much functionality you won't need to update for eternity. Besides once you have it tuned to how you want it, why to you need to change it? Unless you're doing room correction. Either way who offers a better solution? Unless you go PC based with a multichannel DAC.
Yes nothing we can do about innovation making older chips eventually obsolete.
Is this speaker going to be 100% active? Do you plan on just using the DAC's inside the DSP chips? How about maintaining clock sync between the 2 DSP chips?
For the mid/high frequency driver I am hoping to use a passive first order filter, because of the waveguide involved the response falls pretty fast starting at 400 Hz, if the xo is placed at 400 Hz as well it'll probably be a good match for a 24dB slope on the midwoofer, this is my first guess from the measurements I took on the mid/high section, yet to be tested with the woofer. I just need to build some test boxes, I have all the woofers.
So this would be implemented as an active bass / passive mid/high speaker where the user supplies his own DAC and a (hopefully) SET amp for the mids/highs as they will be around 102 dB efficiency. I think for now this is the best way to do things... for my tastes powering a high efficiency mid/high section with a simple passive xo and good tube amps is the way to get a system that is the most engaging and fun to listen to, even if not the most accurate.
For the DSP, probably the only functionality I want apart from xo, eq and driver correction is the ability to adjust the bass to suit the room using either preset curves, a measurement mic, or both. It needs to be kept simple and easy to use.
I'm sure Danville can come up with the perfect solution. I'm not sure a fancy buffer will be essential for 400hz down. Hypex's standard LM4562 buffer would probably work great.
So I assume there is no silly licence manager stuff that nails it down to a specific PC (or OS version)? Do you know if the tools run under an emulated Windows environment, or do they need native Windows?
Are you ever 100% happy with your crossover settings?
So I assume there is no silly licence manager stuff that nails it down to a specific PC (or OS version)? Do you know if the tools run under an emulated Windows environment, or do they need native Windows?
Are you ever 100% happy with your crossover settings?
I'm not sure how to configure their stuff. Probably depends on the board. They have a dizzying about of DSP options. I've had my eye on that new 589 SHARC chip though. Best one yet by far. But I still say PC based DSP is the holy grail.
I don't see too many people ripping apart their passive Xovers to mod for every mood their in. Set them up good for the room your in, and leave them. If you move rooms, maybe tweak a bit more.
To some degree the whole idea of a "PC" is getting very vague. Originally PC meant "Personal computer". Now we all have far too many "personal computers", so PC really means "desktop computer". No, I don't want part of my music system to sit on my desktop and have a full-size screen (I do want to use the computer on my desk, with the full-size screen, to be able to be used as an user interface for my music system, but that is a slightly different story). I do see that general-purpose embeddable computers (such as the Raspberry Pi) are quickly catching up with dedicated DSP processors in terms of capability.
"Set them up and leave them" is pretty much what you had to do with passive crossovers. The benefit of an active, software-configured crossover is that it is easy to modify and experiment - again, that DIY thing...
You simply use the computer's Intel processors to run a powerful DSP engine. Way beyond the power of any DSP chip, and adds no noise or jitter to the signal path. Easily upgradable as well. There's a reason Merging ditched the DSP chip based boards for Masscore.
They use a real time OS that runs the DSP in the background of a Windows installation on its on dedicated cores. Performance is beyond where DSP chips will be in 10 years. Moore's law moves the power of Intel CPU's ahead much faster than DSP chips as well.
Pair with one of their Ravenna multichannel DAC's and there's a real active system
http://www.merging.com/products/pyramix/masscore
Settled on an older laptop due to price reasons, but i was considering a fanless core i5 HTPC as music source, player, DSP. This feeds an 8CH USB DAC --> 8ch ncore,ucd400 amps --> 4way speakers.
PCs do have the overhead of OS, also crashes, virus attack etc are possibilities.
However need for a dedicated screen can be avoided, remote desktop and remote control apps for Jriver etc are available for iphone/ipad/android
Though lower in DSP horsepower, true embedded DSP does get rid of above hassles.
PCs do have the overhead of OS, also crashes, virus attack etc are possibilities.
However need for a dedicated screen can be avoided, remote desktop and remote control apps for Jriver etc are available for iphone/ipad/android
Though lower in DSP horsepower, true embedded DSP does get rid of above hassles.
hypex ncore
You don't want to use the same PC you do your daily internet browsing on for this task. You install a clean OS, and dedicate it for DSP duties. You can run the GUI via remote mobile apps like Jremote, Roon etc.
Embedded chips also add noise, jitter, and are limited in resolution. Try finding a DSP chip that can preform DSP to a 1 bit Native DSD signal, without conversion to PCM.
You don't want to use the same PC you do your daily internet browsing on for this task. You install a clean OS, and dedicate it for DSP duties. You can run the GUI via remote mobile apps like Jremote, Roon etc.
Embedded chips also add noise, jitter, and are limited in resolution. Try finding a DSP chip that can preform DSP to a 1 bit Native DSD signal, without conversion to PCM.
Yes, the platform is dedicated for audio, with a tweaked down OS installation. Only extent of internet access is for streaming audio.
Yes for DIY folks, a PC based active loudspeaker is most flexible and capable.
For an OEM trying to make an active loudspeaker product, embedded DSP probably makes more sense
Yes for DIY folks, a PC based active loudspeaker is most flexible and capable.
For an OEM trying to make an active loudspeaker product, embedded DSP probably makes more sense
- Status
- Not open for further replies.